/
test_neuron_gradient.py
143 lines (125 loc) · 5.17 KB
/
test_neuron_gradient.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
#!/usr/bin/env python3
import unittest
import torch
from captum.attr._core.saliency import Saliency
from captum.attr._core.neuron.neuron_gradient import NeuronGradient
from captum.attr._utils.gradient import _forward_layer_eval
from ..helpers.basic_models import (
BasicModel_ConvNet,
BasicModel_MultiLayer,
BasicModel_MultiLayer_MultiInput,
)
from ..helpers.utils import assertArraysAlmostEqual, BaseTest
class Test(BaseTest):
def test_simple_gradient_input_linear2(self):
net = BasicModel_MultiLayer()
inp = torch.tensor([[0.0, 100.0, 0.0]], requires_grad=True)
self._gradient_input_test_assert(net, net.linear2, inp, (0,), [4.0, 4.0, 4.0])
def test_simple_gradient_input_linear1(self):
net = BasicModel_MultiLayer()
inp = torch.tensor([[0.0, 100.0, 0.0]])
self._gradient_input_test_assert(net, net.linear1, inp, (0,), [1.0, 1.0, 1.0])
def test_simple_gradient_input_relu_inplace(self):
net = BasicModel_MultiLayer(inplace=True)
inp = torch.tensor([[0.0, 5.0, 4.0]])
self._gradient_input_test_assert(
net, net.relu, inp, (0,), [1.0, 1.0, 1.0], attribute_to_neuron_input=True
)
def test_simple_gradient_input_linear1_inplace(self):
net = BasicModel_MultiLayer(inplace=True)
inp = torch.tensor([[0.0, 5.0, 4.0]])
self._gradient_input_test_assert(net, net.linear1, inp, (0,), [1.0, 1.0, 1.0])
def test_simple_gradient_input_relu(self):
net = BasicModel_MultiLayer()
inp = torch.tensor([[0.0, 5.0, 4.0]], requires_grad=True)
self._gradient_input_test_assert(net, net.relu, inp, 0, [0.0, 0.0, 0.0])
def test_simple_gradient_input_relu2(self):
net = BasicModel_MultiLayer()
inp = torch.tensor([[0.0, 5.0, 4.0]])
self._gradient_input_test_assert(net, net.relu, inp, 1, [1.0, 1.0, 1.0])
def test_simple_gradient_multi_input_linear2(self):
net = BasicModel_MultiLayer_MultiInput()
inp1 = torch.tensor([[0.0, 100.0, 0.0]])
inp2 = torch.tensor([[0.0, 100.0, 0.0]])
inp3 = torch.tensor([[0.0, 100.0, 0.0]])
self._gradient_input_test_assert(
net,
net.model.linear2,
(inp1, inp2, inp3),
(0,),
([12.0, 12.0, 12.0], [12.0, 12.0, 12.0], [12.0, 12.0, 12.0]),
(3,),
)
def test_simple_gradient_multi_input_linear1(self):
net = BasicModel_MultiLayer_MultiInput()
inp1 = torch.tensor([[0.0, 100.0, 0.0]])
inp2 = torch.tensor([[0.0, 100.0, 0.0]])
inp3 = torch.tensor([[0.0, 100.0, 0.0]])
self._gradient_input_test_assert(
net,
net.model.linear1,
(inp1, inp2),
(0,),
([5.0, 5.0, 5.0], [5.0, 5.0, 5.0]),
(inp3, 5),
)
def test_matching_output_gradient(self):
net = BasicModel_ConvNet()
inp = torch.randn(2, 1, 10, 10, requires_grad=True)
self._gradient_matching_test_assert(net, net.softmax, inp)
def test_matching_intermediate_gradient(self):
net = BasicModel_ConvNet()
inp = torch.randn(3, 1, 10, 10)
self._gradient_matching_test_assert(net, net.relu2, inp)
def _gradient_input_test_assert(
self,
model,
target_layer,
test_input,
test_neuron,
expected_input_gradient,
additional_input=None,
attribute_to_neuron_input=False,
):
grad = NeuronGradient(model, target_layer)
attributions = grad.attribute(
test_input,
test_neuron,
additional_forward_args=additional_input,
attribute_to_neuron_input=attribute_to_neuron_input,
)
if isinstance(expected_input_gradient, tuple):
for i in range(len(expected_input_gradient)):
assertArraysAlmostEqual(
attributions[i].squeeze(0).tolist(),
expected_input_gradient[i],
delta=0.1,
)
else:
assertArraysAlmostEqual(
attributions.squeeze(0).tolist(), expected_input_gradient, delta=0.1
)
def _gradient_matching_test_assert(self, model, output_layer, test_input):
out = _forward_layer_eval(model, test_input, output_layer)
gradient_attrib = NeuronGradient(model, output_layer)
for i in range(out.shape[1]):
neuron = (i,)
while len(neuron) < len(out.shape) - 1:
neuron = neuron + (0,)
input_attrib = Saliency(
lambda x: _forward_layer_eval(model, x, output_layer)[
(slice(None), *neuron)
]
)
sal_vals = input_attrib.attribute(test_input, abs=False)
grad_vals = gradient_attrib.attribute(test_input, neuron)
# Verify matching sizes
self.assertEqual(grad_vals.shape, sal_vals.shape)
self.assertEqual(grad_vals.shape, test_input.shape)
assertArraysAlmostEqual(
sal_vals.reshape(-1).tolist(),
grad_vals.reshape(-1).tolist(),
delta=0.001,
)
if __name__ == "__main__":
unittest.main()